Consumer Durables and the Optimality of Usually Doing Nothing

Consumer Durables and the Optimality of Usually Doing Nothing

NBER WORKING PAPER SERIES CONSUMER DURABLES AND THE OPTIMALITY OF USUALLY DOING NOTHING Avner Bar-lien Alan S. Blinder Working Paper No. 2488 NATIONAL BUREAU OF ECONOMIC RESEARCH 1050 Massachusetts Avenue Cambridge, MA 02138 January 1988 We thank Andrew Abel, Ben Bernanke, and Sanford Grossman for useful comments, John Amer for research assistance, and the NSF for financial support. The research reported here is part of the NBER's research program in Economic Fluctuations. Any opinions expressed are those of the authors and not those of the National Bureau of Economic Research. Support from The Lynde and Harry Bradley Foundation is gratefully acknowledged. NBER Working Paper #2488 January 1988 Consumer Durables and the Optimality of Usually Doing Nothing ABSTRACT often This paper develops a simple but important point which is overlooked: It is quite possible that the best policy for a rational, optimizing agent is to do nothing for long periods of time—-even if new, relevant information becomes available. We illustrate this point using the market for durable goods. Lumpy costs in durables transactions lead consumers to choose a finite range, not just a single level, for their durables consumption. The boundaries of this range change with new information and, in general, obey the permanent income hypothesis. However, as long as the durable stock is within the chosen region, the consumer will not change her stock. Hence individuals will make durable transactions from the infrequently and their consumption can differ substantially prediction of the strict PIH. Such microeconomic behavior means that aggregate data cannot be generated by a representativeagent; explicit aggregation is required. By doing that, we showed that time series of durable expenditures should be divided to two separate series: One on the average expenditure per purchase and the other on the number of transactions. The predictions of the PIH hold for the former, but not for the latter. For example, the short-run elasticity of the than one number of purchases with respect to permanent income is much larger of for plausible parameter values. We put our theory to a battery empirical tests. Although the tests are by no means always consistent with the theory, most empirical results are in line with our predictions. Avner Bar-Ilan Alan S. Blinder Department of Economics Department of Economics Tel-Aviv University Princeton University Ramat-Aviv, ISRAEL Princeton, NJ 08544 The assumption of optimizing behavior Is the bedrock on which most of economic theory stands. It is also the characteristic that, for better or for sciences. Economists worse, distinguishes economics from the other social -- are probably the only people on earth who believe or act as if they believe -— that homo sapj behaves like homo economicus. This paper does not question the hypothesis that people optimize. it tries to the (Probably, some other paper should.) Instead, square that human assumption of optimizing behavior with the coninon observation behavior seems highly inertial. Like Newton's bodies at rest or in motion, people often seem to cling to their past behavior despite clear evidence costs of calling for change. We will argue that the existence of lumpy changing a decision variable makes inertial behavior optimal under quite idea with a detailed general conditions. And we will illustrate this general analysis of a formal model of the purchase of consumer durables. one which income differs in some significant respects from the standard permanent hypothesis (PIH). 1. RATIONAL INERTIA: BASIC IDEAS The standard type of optimizing behavior posited in economic models is Individual continuous reoptimizatlon. According to this view of behavior, the constraints or firm controlling a decision variable, xt, Is given new and/or the value of the information each period (call that the vector zt), computes decision variable that maximizes his objective function (call it 4), and Is defined some then sets xt=4 in every period. This optimum normally by is kind of tangency condition. Examples abound. In consumption theory, x' current income and wealth, and consumption, zt Includes interest rates, notion expectations about the future; the "Euler equation" Is based on the 1 that the consumer optimizes intertemporally in both period t and period t÷1. In portfolio theory, xt is a vector of portfolio shares and zt includes current wealth and the presumed varlance—covariance matrix; wealthhoiders are supposed to adjust their portfolios every time their perceived variance—covariance matrix changes. Figure 1 is a trivially simple graph displaying this general idea. Here V(x;z) is the objective function, which moves any time z changes. The decision maker Is assumed always to select x = argmax V(x;z) (point E), which induces a behavioral relation of the form xt = F(zt), Changes in z induce prompt responses In xt. —— It seems doubtful, however, that many people behave this way especially If the period is relatively short. (The authors of this paper certainly don't.) Instead, people are alleged to be 'creatures of habit." Econometric evidence certainly supports the general idea that behavior is = inertial. For example, empirically estimated decision rules of the form x F(Zt) are almost always Improved by the addition of xt_1 to the righthand side. Pervasive inertia seems to be a stylized fact of economic life. Inertia may be irrational, as other social scientists will argue. —— Laziness, procrastination, and other human frailties in general, a failure to pursue one's best interests —— can all lead to inertial behavior. About —— this, economists have little to say. But it Is also possible and this Is —— the central point of the paper that Inertia can be rational when there are lumpy costs of changing one's decision variable. It is easy to see why. Ignoring, for the moment, how such costs might arise, suppose that our prototypical decisionmaker incurs a fixed cost, b, each time he changes his control variable. Now look back at Figure 1 and suppose that, because zt has himself at a changed since the last decision period, the decislonmaker finds 2 E P I I I I I I I I II I V ( x ;z) $ II I 'I I II I I I 1 II K s xt_i xSt Fiture 1 Oontinuos Reoptimizatlon versus the (3,s) Rule point like P rather than at point E. Here x is too low, But if V(x) — V(x_1) Is 'small relative to b, It will not pay to raise x. Hence we have argued intuitively that fixed costs will lead to some range around x (shown in the diagram as (s,a)) within which it does not pay to change the decision variable, Within this range, behavior is strictly Inertial: the decisioninaker "does nothing." We like to think of this conclusion as showing that what Akerlof and Yellen (1985) called near rationality is actually full rationality. In the presence of fixed transactions costs, continuous reoptimization would be Irrational. While literally fixed costs are the easiest case to understand, they ar€ not needed to rationalize inertial behavior. Any type of lumpy transactions cost will do. Such costs, we would argue, are pervasive facts of economic life. In some contexts, there are explicit transactions costs —— such as the large commission a real-estate agent receives for selling your house, or the costs a firm Incurs In printing and distributing new price lists ("menu costs"). These costs probably have much to do with why people change houses Infrequently,1 why Investors do not reoptimize their stock portfolios every morning,2 and why firms do not change prices every time either demand or cost changes .3 In other cases, the costs may be implicit. Sometimes the cost Is a time cost, as when consumers spend hours searching for Information on performance —- characterislcs and prices of heterogeneous durable goods not to mention ci houses or jobs. Other times, the transactions cost Is more naturally expressed as a utility cost, as when the individual finds the process of and change or acquiring Information inherently distasteful. (Job Interviews moving may be good examples.) Or the Implicit cost may simply arise from the fact that a human being can only cope with so many problems at once, a point 3 often stressed by Herbert Simon. Hence there is a kind of "shadow cost" for using the scarce resource of mental capacity. (Why else would CEOs delegate decisions?> Finally, some markets have large spreads between the prices at which you can buy and sell, so that a buyer loses a lump of value the moment he takes the good home. This type of lumpy cost is particularly prevalent in the markets for consumer durables. One rationalization was provided by Akerlof (1970), who argued that markets in which goods are heterogeneous and quality is not readily observable are subject to the "lemons principle." (If this is a good car, why are you selling It?) Whether or not caused by the lemons principle, the gap between buying and selling prices may be a major reason why any one consumer is active in the market for any particular durable only sporadically. The rest of the time, he Is "doing nothing." Of course, the econometric fact that xt_1 is almost always a significant determinant of xt has been noti' 1 many times before. The usual "explanation" is the partial-adjustmentmodel, according to which convex costs of changing x make it optimal to adjust xt to x gradually. Under the assumption that adjustment costs are quadratic, a linear partial-adjustment rule like: * xt - xt_1 = x(xt - xt_1) can be derived from rigorous microfoundations. And so the assumption is frequently made, usually without thinking twice about what it means. There are two major problems with this common approach (and many others in particular applications).

View Full Text

Details

  • File Type
    pdf
  • Upload Time
    -
  • Content Languages
    English
  • Upload User
    Anonymous/Not logged-in
  • File Pages
    52 Page
  • File Size
    -

Download

Channel Download Status
Express Download Enable

Copyright

We respect the copyrights and intellectual property rights of all users. All uploaded documents are either original works of the uploader or authorized works of the rightful owners.

  • Not to be reproduced or distributed without explicit permission.
  • Not used for commercial purposes outside of approved use cases.
  • Not used to infringe on the rights of the original creators.
  • If you believe any content infringes your copyright, please contact us immediately.

Support

For help with questions, suggestions, or problems, please contact us